Security agreed for your setup.

These are the safeguards we agree and verify for each agent deployment. Available controls depend on the selected services and contract. Website and booking data are covered by our Privacy notice.

  1. Hosting locations

    We agree hosting regions and verify the selected services before connecting your business data.

  2. Separate company environments

    Company-specific permissions and separation are defined and tested for the agreed deployment.

  3. Audited AWS services

    We review the assurance scope of selected AWS services. Provider certifications do not certify a custom agent deployment.

  4. Processing locations

    We confirm model-processing locations and any international transfers for the selected provider and model.

AWS strengthens its defenses through advanced security research with Anthropic and OpenAI. Project Glasswing · Daybreak

Your data stays yours.

Your business information is used to help you work, with encryption and clear limits on how inference data is handled.

  • Model-training controls

    Before use, we verify the model provider’s terms and settings for excluding business inputs and outputs from model training.

  • Retention limits

    We document retention for inference, logs, conversations and backups separately. Zero retention is promised only where contractually supported and verified.

  • Encrypted at rest and in transit

    We verify encryption and access controls across the services included in your agreed setup.

Access under your control.

  • Passkeys and MFA

    Protect account access with passkeys or multi-factor authentication.

  • Company SSO

    Sign in through your existing company identity provider.

  • Scoped permissions

    Agents can access only the tools and information you authorise.

  • Human approval

    You review sensitive actions before your agent carries them out.